Development of Hybrid Intelligent Models for Prediction Machining Performance Measure in End Milling of Ti6Al4V Alloy with PVD Coated Tool under Dry Cutting Conditions

Salah Salman Al-Zubaidi, Jaharah Abd GHANI, Che Hassan Che Haron, M. N. Mohammed, Adnan N. Jameel Al-Tamimi, Samaher Mohammed Sarhan, Mohd Shukor Salleh, M. Abdulrazaq, Oday Ibraheem Abdullah

2022

Full text Open access: Gold Metadata (CSL-JSON)
DOI 10.3390/lubricants10100236
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

Ti6Al4V alloy is widely used in aerospace and medical applications. It is classified as a difficult to machine material due to its low thermal conductivity and high chemical reactivity. In this study, hybrid intelligent models have been developed to predict surface roughness when end milling Ti6Al4V alloy with a Physical Vapor Deposition PVD coated tool under dry cutting conditions. Back propagation neural network (BPNN) has been hybridized with two heuristic optimization techniques, namely: gravitational search algorithm (GSA) and genetic algorithm (GA). Taguchi method was used with an L27 orthogonal array to generate 27 experiment runs. Design expert software was used to do analysis of variances (ANOVA). The experimental data were divided randomly into three subsets for training, validation, and testing the developed hybrid intelligent model. ANOVA results revealed that feed rate is highly affected by the surface roughness followed by the depth of cut. One-way ANOVA, including a Post-Hoc test, was used to evaluate the performance of three developed models. The hybrid model of Artificial Neural Network-Gravitational Search Algorithm (ANN-GSA) has outperformed Artificial Neural Network (ANN) and Artificial Neural Network-Genetic Algorithm (ANN-GA) models. ANN-GSA achieved minimum testing mean square error of 7.41 × 10−13 and a maximum R-value of 1. Further, its convergence speed was faster than ANN-GA. GSA proved its ability to improve the performance of BPNN, which suffers from local minima problems.